The equivalent length method, the fitting table values, alarm valve and backflow additions, the most common tee-counting error, and a worked example end to end.
Reviewing the OH3 hydraulic report for a logistics warehouse, the required pressure at the pump discharge showed as 6.1 bar; our independent check gave 6.9 bar. The 0.8 bar difference had one cause: the contractor's software had calculated straight pipe friction only and omitted fittings and valves. This article covers the equivalent length method, the items most often skipped on site, and a typical worked example.
What the equivalent length method is
EN 12845 calculates pipe friction loss with the Hazen-Williams formula, in which L is not straight pipe alone but "the equivalent length of pipe plus fittings". The standard makes the point precisely: every fitting where flow changes direction by 45° or more, and every valve, produces friction loss and must be included. So a 22.5° gentle offset produces no fitting loss; but a 45° elbow, a 90° elbow, the branch of a tee, a butterfly valve, an alarm valve — all count.
Equivalent length comes from one of two sources:
- Manufacturer's data — the primary source, and mandatory for special devices such as backflow preventers, strainers and flow indicators.
- The standard's table — used where no manufacturer data exists, calibrated for C = 120 steel pipe.
The values used most on site
The key items from DN20 to DN200, in metres, valid for C = 120; other C values need the tabulated multiplier.
| Fitting / valve | DN25 | DN50 | DN80 | DN100 | DN150 | DN200 |
|---|---|---|---|---|---|---|
| 90° screwed elbow (standard) | 0.77 | 1.5 | 2.4 | 3.0 | 4.3 | 5.7 |
| 90° welded elbow (r/d = 1.5) | 0.36 | 0.69 | 1.1 | 1.4 | 2.0 | 2.6 |
| 45° screwed elbow | 0.40 | 0.76 | 1.3 | 1.6 | 2.3 | 3.1 |
| Standard tee or cross (flow through branch) | 1.5 | 2.9 | 4.8 | 6.1 | 8.6 | 11.0 |
| Gate valve — straight through | — | 0.38 | 0.63 | 0.81 | 1.1 | 1.5 |
| Alarm or check valve — swing type | — | 2.4 | 3.9 | 5.1 | 7.2 | 9.4 |
| Alarm or check valve — mushroom type | — | 12.0 | 19.7 | 25.0 | 35.0 | 47.0 |
| Butterfly valve | — | 2.2 | 3.6 | 4.6 | 6.4 | 8.6 |
| Globe valve | — | 16.0 | 26.0 | 34.0 | 48.0 | 64.0 |
A few practical observations:
- Screwed versus welded long-radius elbows — 3.0 m against 1.4 m at DN100, less than half. On risers and large branches, choosing welded eases the hydraulics noticeably.
- The mushroom check valve penalty — 25.0 m at DN100. A single valve behaves like 25 metres of straight pipe. Two in series near the pump adds a phantom 50 m.
- Globe valves — 16.0 m at DN50. You should not meet one in a sprinkler line; if you do, ask why it is there.
Where both direction and size change
The standard sets a clear rule: where flow direction and pipe size change at the same elbow, tee or cross, equivalent length and pressure loss are calculated on the smaller size. So a 90° elbow sitting exactly at a DN100-to-DN80 transition uses the DN80 value (1.1 m welded), not the DN100 one. That small detail changes the direction of the calculation at large size transitions.
Backflow preventers, alarm valves and special devices
Items not in the table use manufacturer's data. Common additions:
- Backflow preventer (RPZ or double check): the manufacturer gives pressure loss at nominal flow, or a Kv value. A DN100 RPZ typically loses 0.5–0.9 bar at around 6 m³/h, added directly to the total rather than converted to equivalent length.
- Wet alarm valve: the swing-type table values may be used, but most certified manufacturers declare their own figure in their approval documentation — use that.
- Flow indicator (paddle type): a small contribution, typically 0.1–0.3 bar from the manufacturer. Not in the table, so the manufacturer's value binds.
- Strainer (Y-type): very low loss when clean, but an uncleaned strainer can exceed 1 bar. Enter the manufacturer's clean figure and note the maintenance interval in the report.
Common error: counting the tee wrongly
The three errors I meet most often:
- Counting the tee on both legs. The "flow through branch" figure applies only to the branch calculation. The straight-through leg has no fitting contribution, because flow does not change direction. Sometimes the reverse happens and it is omitted from the branch. The rule: add it on the branch, not the run.
- Counting the same tee as a branch on two separate legs. Where a tee splits flow, the branch contribution is not added on both legs; the tee's geometry governs, and the branch is one leg.
- Counting reducers and straight couplings as fittings. Only direction changes of 45° or more count; straight reducers and couplings are practically ignored unless the manufacturer says otherwise. Adding them is not conservative, it is wrong.
Worked example — a DN100 distribution line
On an OH3 site, taking the distribution line furthest from the alarm valve up two floors. The configuration:
- Straight pipe: 42.0 m of DN100 steel (C = 120)
- Fittings: four 90° welded elbows (r/d = 1.5), one tee branch flow, one DN100 gate valve, one swing alarm valve
- Backflow preventer: manufacturer's loss at 1500 L/min is 0.42 bar
- Design flow: 1500 L/min
Total equivalent length:
| Item | Qty | Unit Le (m) | Total (m) |
|---|---|---|---|
| Straight DN100 pipe | — | — | 42.00 |
| 90° welded elbow DN100 | 4 | 1.40 | 5.60 |
| Tee DN100 (branch flow) | 1 | 6.10 | 6.10 |
| Gate valve DN100 | 1 | 0.81 | 0.81 |
| Swing alarm valve DN100 | 1 | 5.10 | 5.10 |
| Total equivalent length | 59.61 |
The straight pipe was 42 m; the equivalent length is 59.61 m — 42 % greater. Because Hazen-Williams friction loss is directly proportional to L, ignoring that difference produces a false pressure curve. The backflow preventer's fixed 0.42 bar is then added directly to the pressure balance.
Approximately: at 1500 L/min in DN100 at C = 120, unit friction loss is around 0.041 bar/m. So 59.61 × 0.041 ≈ 2.44 bar of friction, plus 0.42 bar for the backflow preventer, giving 2.86 bar. Calculating on straight pipe alone (42 × 0.041 ≈ 1.72 bar) misses 1.14 bar of fittings, valve and device contribution. That is exactly where the "unexplained 0.8–1.5 bar difference" between projects comes from.
Comparison with NFPA 13
NFPA 13 gives equivalent lengths in its own table on the same logic, but the numbers are not identical. The NFPA table is also based on C = 120 with multipliers for other C values. In practice the two differ by 10–15 %, which is normal given different calibration sources. On mixed projects — NFPA 13 storage plus EN 12845 office protection in one building — decide at the outset which table applies where, and never mix both within one hydraulic branch.
Turkish practice
BYKHY refers sprinkler hydraulics to TS EN 12845 or NFPA 13 without a separate provision for equivalent length. In practice, fire service approval expects the hydraulic report to show the fittings and valve equivalent lengths as a separate column. The standard's reporting requirement asks for the number, type and equivalent length in metres of the fittings; a report missing that column is returned at inspection.
Quick check list
- Is the fittings breakdown shown in a separate column for every hydraulic branch?
- Is the "flow through branch" value applied only to the branch leg of each tee?
- Is the smaller size used at an elbow where the diameter also changes?
- Are the backflow preventer and alarm valve figures referenced to the manufacturer's data?
- Has the C-value multiplier been applied for galvanised or aged pipe?
- Where a mushroom-type check valve exists, has the correct type been selected rather than the swing value?
Frequently asked questions
How does EN 12845 handle fitting losses?
Through the equivalent length method: every valve and every fitting turning flow by 45° or more is converted into an equivalent length of straight pipe. Values come from the standard's table, with manufacturer's data taking precedence.
What is a DN50 standard 90° elbow worth?
1.5 m for a standard screwed elbow, 0.69 m for a welded long-radius elbow, 2.9 m for a tee branch flow, 0.38 m for a gate valve and 2.4 m for a swing-type alarm or check valve.
Is a tee counted on both the inlet and outlet legs?
Consider whether flow changes direction in each leg. The branch leg takes the "flow through branch" value; the straight-through leg takes none. A direction change of less than 45° produces no fitting contribution.
Where does the backflow preventer figure come from?
Backflow preventers, strainers and flow indicators are not in the table. The standard requires manufacturer's data first — a Kv value or a stated equivalent length. Never assume a figure.
Why does a mushroom check valve lose so much more than a swing type?
Because flow must bend around the disc in a mushroom (spring-loaded axial) valve, while a swing disc opens in line with the flow. At DN100 the swing alarm valve is 5.1 m against 25.0 m for the mushroom — roughly five times.

SprinkCalc — Fire Sprinkler Design Across Three Standards
SprinkCalc covers hazard classification, design density and area, K-factor selection, water demand and hydraulic calculations for NFPA 13, FM Global and BS EN 12845 in a single iOS app, and exports a professional PDF report.
Download SprinkCalc on the App Store
MEP Calc — 86+ Engineering Calculators
MEP Calc bundles 86+ engineering modules in one iOS app: 21 fire calculations plus heating, cooling, HVAC, plumbing, steam and natural gas.
Download MEP Calc on the App StoreBS EN 12845:2015+A1:2019 Fixed firefighting systems — Automatic sprinkler systems. EN 12845-2:2024 (CMSA & ESFR sprinkler systems). NFPA 13 Standard for the Installation of Sprinkler Systems. Turkish Regulation on Fire Protection of Buildings (BYKHY). FM Global Property Loss Prevention Data Sheet 2-0.